報告題目：Graph Neural Network: An Introduction and Some Recent Works
報告人：周川，中國科必一數學與系統科學研究院副研究員
報告時間：2021 年 10 月 29 日（星期五）下午 18: 00
騰訊會議：815 741 454
報告摘要: Learning with graph structured data, such as social, biological, and financial networks, requires effective representation of their graph structure. Recently, there has been a surge of interest in Graph Neural Network (GNN) approaches for graph representation learning. GNN generalizes neural network (CNN) from low-dimensional regular grids, where image, video and speech are represented, to graph structure data. To date, GNN has been successfully applied to many noteworthy applications, such as node classification, link prediction, recommendation and traffic prediction. This speech will mainly review the GNN with the background, emerging challenges, basic concepts, state-of-the-art algorithms, and some of our recent works.
報告人簡介：周川，現為中國科必一數學與系統科學研究院副研究員，博士生導師。研究方向為社會計算、社交網絡分析、圖挖掘、統計機器學習等，在國際頂級學術期刊和會議（如TKDE、DMKD、PR、AAAI、IJCAI、ICDM、CIKM等）上累計發表學術論文60余篇。承擔國家自然科學基金、國家重點研發計劃等10余項科研課題。獲得中科院數必一陳景潤未來之星、CCF Senior Member(即中國計算機學會高級會員)等榮譽。